This is right on the money. In fact, if you think about it, the crux of Microsoft's response to the DoJ's antitrust suit is just this--that it is fundamentally impossible (whether that is technically true or not) to remove such add-on pieces from the underlying operating system. Not only is this all-in-one approach emblematic of Microsoft's thinking, but other approaches (eg, base + optional components) would make their antitrust defense ring hollow.
To my eyes, this is one of the core differences between using Linux software and Microsoft software. With Microsoft software, lots of tools, drivers, etc. come in one shiny box (provided you can afford it), and it does everything one could want, and lots of other stuff besides. With Linux software, each additional functional component is optional. As a result, in consumer markets, Microsoft is perceived as "easier," which is often the case for the majority of users. In highly technical markets such as embedded operating systems, Linux has far more to offer, as one is not locked into a monolithic set of components/functionality (licensing issues aside).
I find it hard to believe that making a mockery of yourself on the web does much to "protect the brand." Unless, of course, the brand is based on mindless legalism.
"Let's see, I'm looking for a fast-moving, innovative organization on whom to lavish the extra $10 million in my IT services budget.... I know! How about I give it to a company that protects its brand as a leader in e-business by prohibiting links to its own site? And that commissioned the worst piece of recorded music ever committed to magnetic tape? Now where's my extra-large checkbook...."
I'm not much of a code jockey, but I know a bit about finance, having worked as an investment banker and venture capitalist.
1. Paying a dividend is not required by the IRS, the SEC or any federal law, as far as I know.
2. Paying a dividend exposes your money to double taxation: the standard corporate rate (paid at the time the profit was earned), and then your personal tax rate (including city, state, and federal). This is money out of your pocket as much as Bill's.
3. Whether Bill Gates ownes 25% of MSFT is irrelevant to the decision to pay a dividend. If I own one share, I'd rather receive the full value of my 1/# shares ownership of Microsoft's cash assets, rather than shave off an additional percentage to go to the government. Paying more taxes is not in the interest of Microsoft's shareholders, no matter how large their holdings. If anything, more stock held by management is a good thing--it aligns the incentives of the managers with the incentives of the stockholders. In a company where the management owns very little of the shares, you are far more likely to see decisions made that sell out shareholders (the owners) to the benefit of the current management (with foolhardy acquisitions, for example). This is a classic problem in economics, and it's why you rarely see small, owner-managed companies being as stupid with their money as large companies often are.
4. If Microsoft can't find profitable uses for the cash, it should buy back its stock. That, as one poster wisely pointed out, will increase the value of each share, as the same pie is divided by a smaller number of shares. Besides, the value of the cash on hand increases the book value of a single share of Microsoft--in other words, each share of ownership now entitles you to your proportional share of each additional dollar in MSFT's bank account, a gain in value that cost you nothing.
5. For most stockholders, the capital gains tax rate is lower (often far lower) than the personal income tax rate. As a result, you'd rather get paid back for your investment in Microsoft in stock price appreciation than tax-affected cash.
6. People buy equities (shares of stock) because they represent a future stream of cash flows generated from a business. In this way, buying the stock of a business is no different than buying a bond: both represent a stream of future payments, and the price you pay now depends on the magnitude, frequency, and likelihood of those payments. Thus, buying the stock of a company that doesn't pay dividends is like buying a zero-coupon bond (a bond that doesn't pay periodic interest, but that reaches full face value at maturity). The net result is that market adjusts the price you paid for the stock or bond to account for the nature of the payouts. Instead of valuing the stream of cash payments you receive in the form of dividends, you value the discounted cash flows that Microsoft retains.
But Microsoft shouldn't just sock away money in a money market account. If it can't find opportunities that offer a greater rate of return than a money market account or bonds (both vehicles which are available to individual stockholders as well), then it should buy back stock. (Technically, the rate-of-return hurdle for new lines of business or new acquisitions should be Microsoft's cost of capital, the rate at which it can "borrow" money from the capital markets.)
Other good sites with compatibility information are from Lifeclips.com and YesVideo.com, services which will transfer your videos from other formats (Hi-8, VHS, etc.) to DVD, along with automatic scene detection, chapters, menus, etc.
I don't know the specifics of your option agreement, and I'm not a lawyer (though I am a venture capitalist), but you might be able to preserve your vested options via what is known as an "83b election."
This allows you to exercise your vested options (even if the stock is not yet traded publicly), which means that you own them outright, even if you leave. Also, it allows you to take capital gains tax treatment of any increase in value in your options (as opposed to standard income tax treatment), which can potentially save you money in taxes, depending on your bracket.
However, exercising your options means writing a check for the number of vested options * strike price. That can be a big check, and if the company goes under or never goes public/gets acquired, you may never see that money again. Also, it can have severe tax implications depending on the circumstances surrounding the value of those options depending on what happens to the company.
Before you even THINK of doing something like this, talk to a tax lawyer/accountant first. Also, your company may or may not allow this sort of thing.
I've never done one of these myself, but I know people who have. Just an option. Use this advice at your own risk.
This is from a BusinessWeek article describing the BSA and its "enforcement actions."
Apparently, the number of actions totalled 550 in the year leading up to March 97 (when the article was published), of which only 10% were raids, the BSA claims.
I'm sure there's been a slashdot thread on this already, but I just wanted to mention that Equifax Secure might be a useful solution to those looking for cheaper server certificates (vs. Verisign/Thawte).
They used to be $49, but apparently they've raised their prices to $79. They claim that their certificates will work with Apache+SSLeay and Apache+Raven. I am wondering if anyone has had experience with using Equifax certificates (in general), and specifically whether they work with Apache+mod_ssl?
Also, they offer "wildcard" certificates, which allow you to secure *.yourdomain.tld, which seem pretty interesting for an app I'm working on. Any experience with these?
The new DirecTivo service from DirecTV has built-in 5.1 channel digital audio. It is my understanding that in these new boxes, there is no chip to do MPEG encoding/decoding--it just writes the digital signal directly from the satellite feed (maybe there's some format conversion or something).
This also means that there are no quality settings on the DirecTivo box: it's always "Best" quality, since the signal is always digital. The recording time is still 35 hours (I'm told). Sounds pretty cool. The downside is that you can't record cable or off-air TV into it as well, since there is no MPEG encoding capability.
In 1995, an MIT student named David LaMacchia was prosecuted for allegedly distributing copyrighted software via and FTP server he set up on MIT's Athena workstations.
He was prosecuted by the federal government under federal wire fraud statutes, but the case was dismissed because the judge found that copyright infringement cannot be prosecuted under the wire fraud statute.
I wonder if the DMCA has superseded this precedent (though the Massachusetts case may not apply in Oklahoma), which seemed to make FTP sites into "common carriers" in the eyes of the law. More information here.
For an intranet site that generates signed correspondence and formatted fax cover sheets, I have very successfully used a free HTML to PDF converter from Easy SW, called HTMLDOC.
Use PHP to write out formatted HTML (that's the tweak-heavy part) to disk, then use a shell command to run HTMLDOC and convert it to PDF, and display it in the browser. We've generated literally thousands of documents this way, and it works great. All free as in all free.
Much easier than using the PDFlib library that comes with PHP, as you can avoid having to learn anything about PostScript. You are at the mercy of HTMLDOC's formatting, which can be quirky. But it's vastly improved in the short time we've been using it, and new versions are out almost biweekly.
I'm curious how this is different from Slashdot publishing the infamous "Halloween Memo," which was internal and confidential to (and presumably copyrighted by) Microsoft.
Would it have been illegal to post the contents of that document in its entirety, rather than just linking to it (as was the style at the time)?
How is it that the press can get away with reprinting such documents in their entirety, but AC's and/. posters can't? The Pentagon Papers (which were top secret), the Halloween memo, internal tobacco company documents, etc. are all examples of news organizations publishing restricted material. And the Supreme Court ruled in favor of the NYT in the Pentagon Papers case.
Also, does this mean that as long as I put an installshield wrapper with a EULA around my documents, that I can start to exact DMCA revenge in these cases?
How about website materials? Would I be forbidden from posting an item from the Microsoft Knowledgebase[sic], because it says "copyright MSFT"? Or that's okay because it doesn't have a EULA around it?
What I think is interesting in this case (and interesting about the Net in general) is how it allows for a kind of "general arbitrage."
Consider this analogy: Arbitrage, for the uninitiated, as applied in the stock market means to exploit a pricing inefficiency in the marketplace by making a trade that closes the inefficiency. Arbtriage can be riskless--ie, if a merger is announced at $3 per share, and the stock is trading at $1.50, buying at $2 gives you a riskless return. That's why these differences tend not to exist very long.
The key to arbitrage is availabilty of information--if everyone in the world knows at the same moment that a stock trading at $1.50 should be valued at $3, the price will immediately become $3.
Why this is relevant: What the internet allows is arbitrage over a wide range of "markets" (including the financial markets). Because information can be served or accessed from anywhere, countries are having difficultly enforcing "legal inefficiencies" that stand in the way of what the users want.
For instance: --DeCSS banned in US? Serve it from a box in a country with no extradition treaty with the US --UK ISPs remove content at the hint of a threat? Host it in the US, where laws offer better protection to ISPs. --Gambling illegal in US? Host it from the Bahamas. --Pricing discrepancy between Buy.com price for a 19in Monitor and retail store pricing? 800 orders. --No porn in Australia? A data center in Santa Clara, CA, is the new home for Aussie porn.
Now maybe this is entirely self-evident, based on the design of the internet. And sure, there are ways that nations or companies could combat this, should they really want to.
But as long as there is one country with relatively open interpretations of laws that are strict in other countries, and as long as the technology of the internet continues to allow users to access that "forbidden" content relatively easily, this phenomenon seems likely to continue. Of course, governments could decide to inspect every inbound and outbound packet from their country--but it's not easy.
Maybe instead of tax havens, there will be "Net Havens." Some might call this a "race to the bottom," but I think on balance more people will end up with access to the expression/commerce they desire. I think that's a good thing. And the countries that tend to be more lax in enforcing restrictions on speech/commerce will likely benefit from more internet traffic/hosting business.
Okay, somehow the domain I thought I was registering got hijacked by someone else. Very strange. That may merit a separate investigation. (I went to a decent, linux-friendly ISP a friend had used, filled out the form for a new registration, and the new domain appeared, but under someone else's name--someone else who evidently sells domain names. Is this common?) I can't imagine that someone reading this thread went out to register the domain and beat me to it. That's just not cool, and it's not something I would expect from the/. crowd.
Anyway, I don't have the domain I thought I had. I still plan to put something together on this topic, so contact me at vcinfo@yahoo.com for the site. (I think this thread will be archived before I can get a new domain ready.)
I have started the process of getting a site up. I registered the domain yesterday, but I don't think it's propagated yet--I can't get to it myself. There certainly isn't any content yet.
BTW, I am doing this independently of my firm, so I had a developer friend register the domain for me (ie, I'm not the guy in the whois database)--I just don't want anyone to get the idea that I speak for my firm or its partners, because I don't.
The tentative address is vcinfo.com. While I'm moderately technical, this is my first domain registration, so there's a non-zero chance I screwed it up. I have yet to set up an email address at that domain, so for the moment, send ideas for the site to vcinfo@yahoo.com.
Believe it or not, some VCs do read slashdot. I am an analyst/associate at a prominent, early-stage VC firm on Sand Hill Road in Silicon Valley, with several hundred million dollars under management. We have invested in many technology companies, and we have seen but rejected orders of magnitude more (that's the nature of the business).
Some clarifications: First of all, we don't sign NDAs. Neither do any other reputable VCs along Sand Hill Road, because we'd quickly be swamped with them (we review upwards of 10,000 plans a year). Our business depends on our reputation, and nothing else--everyone's money is green, and our competitors are literally everywhere along Sand Hill. If someone approaches us with an NDA, and refuses to talk until we sign, we pass on the deal.
Second, VCs do take a big ownership stake. But it's not just about money. We have the connections into our portfolio companies and into other large companies around the Valley and beyond. As an early stage firm, we work for our money. We help recruit talent, we help with later financings, we help make strategic decisions, etc. CEOs often talk with their lead investors several times a week.
Third, regarding patents, they're nice, but what's more important is defensibility in the marketplace. With the exception of foundational algorithms like RSA, it's unlikely that patent protection will be all that significant. It's more important to have a lead over potential competition due to a combination of better technology, speed, and strategy.
Finally, no bank in the world is likely to give a loan to a startup company without financial backing--you're simply too risky, and there's no assets to recover. VCs can handle the risk; we're used to it.
As far as advice goes: 1. Make progress. Prove your assumptions--people want your product, say they will pay for your product, want an alliance with you. Move your technology as far along as possible. Hire good people.
If you have had no professional money invested yet, consider approaching angel investors Garage.com, Angel Investors LP. They will help you advance to the stage where VC money is appropriate. If it's two guys and an idea, we generally won't back it unless it's people we know and trust already. And sometimes not even then.
2. Think about the strategic landscape of the market you're entering now, 6 mos from now, a year from now, and beyond. What will happen once you enter the market? Most importantly, what problem are you solving, and who will pay you to solve it? How much will they pay? How will you market to/sell to them? VCs will ask tough questions about everything--you, your competition, your technology, your strategy, etc. Good VCs aren't just money managers--they are engineers with MBAs. Many have CS or EE degrees, and all are up on the latest technologies. You'd be surprised how many technical discussions take place here.
3. Do your homework. Figure out who else is out there, what they're doing, and make yourself stand out. Be the diamond in the rough of 10000 plans we see. For advice on how to do that, check out MIT's 50K page. Pay special attention to the "resources" section.
There are dozens of other pieces of advice to offer. So I am considering creating a website with this and other information on the VC process. Does that seem like something people would be interested in? If so, please post a reply, and I'll start putting one together.
I may not know much about the kernel, but finally a subject I can be useful on!
Slashdot Mirror
This is right on the money. In fact, if you think about it, the crux of Microsoft's response to the DoJ's antitrust suit is just this--that it is fundamentally impossible (whether that is technically true or not) to remove such add-on pieces from the underlying operating system. Not only is this all-in-one approach emblematic of Microsoft's thinking, but other approaches (eg, base + optional components) would make their antitrust defense ring hollow.
To my eyes, this is one of the core differences between using Linux software and Microsoft software. With Microsoft software, lots of tools, drivers, etc. come in one shiny box (provided you can afford it), and it does everything one could want, and lots of other stuff besides. With Linux software, each additional functional component is optional. As a result, in consumer markets, Microsoft is perceived as "easier," which is often the case for the majority of users. In highly technical markets such as embedded operating systems, Linux has far more to offer, as one is not locked into a monolithic set of components/functionality (licensing issues aside).
I find it hard to believe that making a mockery of yourself on the web does much to "protect the brand." Unless, of course, the brand is based on mindless legalism.
"Let's see, I'm looking for a fast-moving, innovative organization on whom to lavish the extra $10 million in my IT services budget.... I know! How about I give it to a company that protects its brand as a leader in e-business by prohibiting links to its own site? And that commissioned the worst piece of recorded music ever committed to magnetic tape? Now where's my extra-large checkbook...."
I'm not much of a code jockey, but I know a bit about finance, having worked as an investment banker and venture capitalist.
1. Paying a dividend is not required by the IRS, the SEC or any federal law, as far as I know.
2. Paying a dividend exposes your money to double taxation: the standard corporate rate (paid at the time the profit was earned), and then your personal tax rate (including city, state, and federal). This is money out of your pocket as much as Bill's.
3. Whether Bill Gates ownes 25% of MSFT is irrelevant to the decision to pay a dividend. If I own one share, I'd rather receive the full value of my 1/# shares ownership of Microsoft's cash assets, rather than shave off an additional percentage to go to the government. Paying more taxes is not in the interest of Microsoft's shareholders, no matter how large their holdings. If anything, more stock held by management is a good thing--it aligns the incentives of the managers with the incentives of the stockholders. In a company where the management owns very little of the shares, you are far more likely to see decisions made that sell out shareholders (the owners) to the benefit of the current management (with foolhardy acquisitions, for example). This is a classic problem in economics, and it's why you rarely see small, owner-managed companies being as stupid with their money as large companies often are.
4. If Microsoft can't find profitable uses for the cash, it should buy back its stock. That, as one poster wisely pointed out, will increase the value of each share, as the same pie is divided by a smaller number of shares. Besides, the value of the cash on hand increases the book value of a single share of Microsoft--in other words, each share of ownership now entitles you to your proportional share of each additional dollar in MSFT's bank account, a gain in value that cost you nothing.
5. For most stockholders, the capital gains tax rate is lower (often far lower) than the personal income tax rate. As a result, you'd rather get paid back for your investment in Microsoft in stock price appreciation than tax-affected cash.
6. People buy equities (shares of stock) because they represent a future stream of cash flows generated from a business. In this way, buying the stock of a business is no different than buying a bond: both represent a stream of future payments, and the price you pay now depends on the magnitude, frequency, and likelihood of those payments. Thus, buying the stock of a company that doesn't pay dividends is like buying a zero-coupon bond (a bond that doesn't pay periodic interest, but that reaches full face value at maturity). The net result is that market adjusts the price you paid for the stock or bond to account for the nature of the payouts. Instead of valuing the stream of cash payments you receive in the form of dividends, you value the discounted cash flows that Microsoft retains.
But Microsoft shouldn't just sock away money in a money market account. If it can't find opportunities that offer a greater rate of return than a money market account or bonds (both vehicles which are available to individual stockholders as well), then it should buy back stock. (Technically, the rate-of-return hurdle for new lines of business or new acquisitions should be Microsoft's cost of capital, the rate at which it can "borrow" money from the capital markets.)
I hope that helps.
Other good sites with compatibility information are from Lifeclips.com and YesVideo.com, services which will transfer your videos from other formats (Hi-8, VHS, etc.) to DVD, along with automatic scene detection, chapters, menus, etc.
Their compatibility lists are here:
Lifeclips Compatible DVD Players
YesVideo Compatible DVD Players
As far as I know, these companies use standard PC-based DVD burners, but I could be mistaken.
This allows you to exercise your vested options (even if the stock is not yet traded publicly), which means that you own them outright, even if you leave. Also, it allows you to take capital gains tax treatment of any increase in value in your options (as opposed to standard income tax treatment), which can potentially save you money in taxes, depending on your bracket.
However, exercising your options means writing a check for the number of vested options * strike price. That can be a big check, and if the company goes under or never goes public/gets acquired, you may never see that money again. Also, it can have severe tax implications depending on the circumstances surrounding the value of those options depending on what happens to the company.
Before you even THINK of doing something like this, talk to a tax lawyer/accountant first. Also, your company may or may not allow this sort of thing.
I've never done one of these myself, but I know people who have. Just an option. Use this advice at your own risk.
Apparently, the number of actions totalled 550 in the year leading up to March 97 (when the article was published), of which only 10% were raids, the BSA claims.
Check it out her e, courtesy of Google's cache.
They used to be $49, but apparently they've raised their prices to $79. They claim that their certificates will work with Apache+SSLeay and Apache+Raven. I am wondering if anyone has had experience with using Equifax certificates (in general), and specifically whether they work with Apache+mod_ssl?
Also, they offer "wildcard" certificates, which allow you to secure *.yourdomain.tld, which seem pretty interesting for an app I'm working on. Any experience with these?
This also means that there are no quality settings on the DirecTivo box: it's always "Best" quality, since the signal is always digital. The recording time is still 35 hours (I'm told). Sounds pretty cool. The downside is that you can't record cable or off-air TV into it as well, since there is no MPEG encoding capability.
Check out Tivo's page on it.
The classic tale of the most bad-ass, plastics-hungry space fungus yet. Read the book, skip the movie.
He was prosecuted by the federal government under federal wire fraud statutes, but the case was dismissed because the judge found that copyright infringement cannot be prosecuted under the wire fraud statute.
I wonder if the DMCA has superseded this precedent (though the Massachusetts case may not apply in Oklahoma), which seemed to make FTP sites into "common carriers" in the eyes of the law. More information here.
Use PHP to write out formatted HTML (that's the tweak-heavy part) to disk, then use a shell command to run HTMLDOC and convert it to PDF, and display it in the browser. We've generated literally thousands of documents this way, and it works great. All free as in all free.
Much easier than using the PDFlib library that comes with PHP, as you can avoid having to learn anything about PostScript. You are at the mercy of HTMLDOC's formatting, which can be quirky. But it's vastly improved in the short time we've been using it, and new versions are out almost biweekly.
Would it have been illegal to post the contents of that document in its entirety, rather than just linking to it (as was the style at the time)?
How is it that the press can get away with reprinting such documents in their entirety, but AC's and /. posters can't? The Pentagon Papers (which were top secret), the Halloween memo, internal tobacco company documents, etc. are all examples of news organizations publishing restricted material. And the Supreme Court ruled in favor of the NYT in the Pentagon Papers case.
Also, does this mean that as long as I put an installshield wrapper with a EULA around my documents, that I can start to exact DMCA revenge in these cases?
How about website materials? Would I be forbidden from posting an item from the Microsoft Knowledgebase[sic], because it says "copyright MSFT"? Or that's okay because it doesn't have a EULA around it?
Consider this analogy:
Arbitrage, for the uninitiated, as applied in the stock market means to exploit a pricing inefficiency in the marketplace by making a trade that closes the inefficiency. Arbtriage can be riskless--ie, if a merger is announced at $3 per share, and the stock is trading at $1.50, buying at $2 gives you a riskless return. That's why these differences tend not to exist very long.
The key to arbitrage is availabilty of information--if everyone in the world knows at the same moment that a stock trading at $1.50 should be valued at $3, the price will immediately become $3.
Why this is relevant:
What the internet allows is arbitrage over a wide range of "markets" (including the financial markets). Because information can be served or accessed from anywhere, countries are having difficultly enforcing "legal inefficiencies" that stand in the way of what the users want.
For instance:
--DeCSS banned in US? Serve it from a box in a country with no extradition treaty with the US
--UK ISPs remove content at the hint of a threat? Host it in the US, where laws offer better protection to ISPs.
--Gambling illegal in US? Host it from the Bahamas.
--Pricing discrepancy between Buy.com price for a 19in Monitor and retail store pricing? 800 orders.
--No porn in Australia? A data center in Santa Clara, CA, is the new home for Aussie porn.
Now maybe this is entirely self-evident, based on the design of the internet. And sure, there are ways that nations or companies could combat this, should they really want to.
But as long as there is one country with relatively open interpretations of laws that are strict in other countries, and as long as the technology of the internet continues to allow users to access that "forbidden" content relatively easily, this phenomenon seems likely to continue. Of course, governments could decide to inspect every inbound and outbound packet from their country--but it's not easy.
Maybe instead of tax havens, there will be "Net Havens." Some might call this a "race to the bottom," but I think on balance more people will end up with access to the expression/commerce they desire. I think that's a good thing. And the countries that tend to be more lax in enforcing restrictions on speech/commerce will likely benefit from more internet traffic/hosting business.
Anyway, I don't have the domain I thought I had. I still plan to put something together on this topic, so contact me at vcinfo@yahoo.com for the site. (I think this thread will be archived before I can get a new domain ready.)
BTW, I am doing this independently of my firm, so I had a developer friend register the domain for me (ie, I'm not the guy in the whois database)--I just don't want anyone to get the idea that I speak for my firm or its partners, because I don't.
The tentative address is vcinfo.com. While I'm moderately technical, this is my first domain registration, so there's a non-zero chance I screwed it up. I have yet to set up an email address at that domain, so for the moment, send ideas for the site to vcinfo@yahoo.com.
Content coming soon!
Some clarifications: First of all, we don't sign NDAs. Neither do any other reputable VCs along Sand Hill Road, because we'd quickly be swamped with them (we review upwards of 10,000 plans a year). Our business depends on our reputation, and nothing else--everyone's money is green, and our competitors are literally everywhere along Sand Hill. If someone approaches us with an NDA, and refuses to talk until we sign, we pass on the deal.
Second, VCs do take a big ownership stake. But it's not just about money. We have the connections into our portfolio companies and into other large companies around the Valley and beyond. As an early stage firm, we work for our money. We help recruit talent, we help with later financings, we help make strategic decisions, etc. CEOs often talk with their lead investors several times a week.
Third, regarding patents, they're nice, but what's more important is defensibility in the marketplace. With the exception of foundational algorithms like RSA, it's unlikely that patent protection will be all that significant. It's more important to have a lead over potential competition due to a combination of better technology, speed, and strategy.
Finally, no bank in the world is likely to give a loan to a startup company without financial backing--you're simply too risky, and there's no assets to recover. VCs can handle the risk; we're used to it.
As far as advice goes: 1. Make progress. Prove your assumptions--people want your product, say they will pay for your product, want an alliance with you. Move your technology as far along as possible. Hire good people.
If you have had no professional money invested yet, consider approaching angel investors Garage.com, Angel Investors LP. They will help you advance to the stage where VC money is appropriate. If it's two guys and an idea, we generally won't back it unless it's people we know and trust already. And sometimes not even then.
2. Think about the strategic landscape of the market you're entering now, 6 mos from now, a year from now, and beyond. What will happen once you enter the market? Most importantly, what problem are you solving, and who will pay you to solve it? How much will they pay? How will you market to/sell to them? VCs will ask tough questions about everything--you, your competition, your technology, your strategy, etc.
Good VCs aren't just money managers--they are engineers with MBAs. Many have CS or EE degrees, and all are up on the latest technologies. You'd be surprised how many technical discussions take place here.
3. Do your homework. Figure out who else is out there, what they're doing, and make yourself stand out. Be the diamond in the rough of 10000 plans we see. For advice on how to do that, check out MIT's 50K page. Pay special attention to the "resources" section.
There are dozens of other pieces of advice to offer. So I am considering creating a website with this and other information on the VC process. Does that seem like something people would be interested in? If so, please post a reply, and I'll start putting one together.
I may not know much about the kernel, but finally a subject I can be useful on!